A tokenised pound is only a pound if the protections travel with it

The first payment was never the hard part

Banks have been moving sterling between customers electronically for half a century. What is new in the pilot announced this week is the container: a deposit recorded as a token on a shared ledger, transferable under conditions written in code rather than instructed through a batch file. UK banks say they have collaborated to complete the first live customer transactions using tokenised sterling deposits. That is genuine engineering, and it works.

The engineering was the easy half. The hard half is the sentence nobody has published: after the token moves, who owes the holder the money?

A deposit is a promise, not a substance

Money in a bank account is not a pile of pounds sitting somewhere with your name on it. It is a promise by one specific institution to pay you on demand. That promise is accepted at face value by everyone, without anyone checking the bank’s books, because of scaffolding built up over a century: capital requirements, supervision, resolution powers that let the authorities take a failing bank apart over a weekend, and a compensation scheme that pays eligible depositors up to a statutory limit if all of that fails.

None of that scaffolding attaches to a database format. It attaches to a legal relationship. Tokenising the record of a deposit does not automatically carry the protections across; whether it does depends entirely on the terms of the token, and those terms are the one thing an announcement of a successful transaction does not tell you.

Whose claim is it after the transfer?

Here is why this is not a lawyer’s quibble. The whole point of tokenised deposits is that value can move between customers of different banks, instantly, against simultaneous delivery of an asset or the discharge of a contract. At the moment of that transfer there are only two possibilities.

Either the receiving customer now holds a claim against the sending customer’s bank — an institution they did not choose, whose creditworthiness they never assessed, with a claim that may sit outside any protection limit they thought applied to them. Or the token is redeemed at the issuing bank and re-issued by the receiving bank, with the interbank difference settled in central bank money, in which case the token is a messaging improvement wrapped around the existing plumbing.

Those are radically different products. They carry different credit risk, different failure modes and different regulatory homes. They are currently being described by the same two words. Which one has just gone live with real customers is the first question a supervisor should ask, and the answer belongs in public.

Three words doing a lot of work

“Tokenised” is doing the least work of the three. “Deposit” is doing the most, because it is the word that borrows a century of public trust and statutory protection. “First” is doing the sneakiest work, because in payments, first tends to become default: the redemption rule, the settlement finality convention and the treatment of a token issued by a bank that subsequently fails get chosen once, in a pilot, by engineers with a deadline, and then everything built afterwards inherits them.

This is also the context in which Europe is arguing about whether to embrace privately issued stablecoins. Tokenised deposits are the banking industry’s answer to that argument, and it is a good answer — same money, better plumbing, existing regulatory perimeter. But the answer only works if the second half is true. A tokenised deposit whose protection status is undocumented is not obviously safer than a stablecoin whose reserve composition is disclosed monthly.

The case against this argument

The strongest objection is proportionality. This is a pilot. The sums will be small, the customers will have consented, and the participants are supervised institutions who did not do this without telling anyone. English law characterises financial instruments by their substance rather than the technology used to record them, so if the customer has a claim on a bank repayable on demand, it is a deposit, and protection follows automatically. Demanding a published rulebook before a controlled experiment is exactly the reflex that makes people complain the UK talks about innovation and delivers consultations.

That objection is mostly right, and it still loses. If substance governs, then saying so publicly costs the participants nothing. A paragraph confirming that holders of these tokens are depositors of a named bank, protected on the same terms as any other balance, and describing what happens to an outstanding token if the issuer fails, is not a regulatory burden. It is a product disclosure. The reason it has not appeared is more likely that the question is genuinely unresolved — and if it is unresolved, “live customer transactions” is the wrong stage to still be resolving it.

What would settle it

One document, four facts: which institution owes the holder before and after a transfer; whether the claim is an eligible deposit; how the interbank leg is extinguished and in what money; and the order of events if an issuing bank enters resolution with tokens outstanding. Every participant already knows the answers, or has discovered that they do not.